A novel amino acid metabolism-related gene risk signature for predicting prognosis in clear cell renal cell carcinoma

BACKGROUND: Renal cancer is one of the most lethal cancers because of its atypical symptoms and metastatic potential. The metabolism of amino acids and their derivatives is essential for cancer cell survival and proliferation. Thus, the construction of the amino acid metabolism-related risk signature might enhance the accuracy of the prognostic model and shed light on the treatments of renal cancers. METHODS: RNA expression and clinical data were downloaded from Santa Cruz (UCSC) Xena, GEO, and ArrayExpress databases. The “DESeq2” package identified the differentially expressed genes. Univariate COX analysis selected prognostic genes related to the metabolism of amino acids. Patients were divided into two clusters using the “ConsensusClusterPlus” package, and the CIBERSORT, ESTIMATE methods were explored to assess the immune infiltrations. The LASSO regression analysis constructed a risk model which was evaluated the prediction accuracy in two independent cohorts. The genomic alterations and drug sensitivity of 18-LASSO-genes were assessed. The differentially expressed genes between two clusters were used to perform functional enrichment analysis and weighted gene co-expression network analysis (WGCNA). Furthermore, external validation of TMEM72 expression was conducted in the FUSCC cohort containing 33 ccRCC patients. RESULTS: The amino acid metabolism-related genes had significant correlations with prognosis. The patients in Cluster A demonstrated better survival, lower Treg cell proportion, higher ESTIMATE scores, and higher cuproptosis-related gene expressions. Amino acid metabolism-related genes with prognostic values were used to construct a risk model and patients in the low risk group were associated with improved outcomes. The Area Under Curve of the risk model was 0.801, 0.777, and 0.767 at the first, second, and third year respectively. The external validation cohort confirmed the stable prognostic value of the risk model. WGCNA identified four gene modules correlated with immune cell infiltrations and cuproptosis. We found that TMEM72 was downregulated in tumors by using TCGA, GEO datasets (p<0.001) and the FUSCC cohort (p=0.002). CONCLUSION: Our study firstly constructed an 18 amino acid metabolism related signature to predict the prognosis in clear cell renal cell carcinoma. We also identified four potential gene modules potentially correlated with cuproptosis and identified TMEM72 downregulation in ccRCC which deserved further studies.